1 / 0

Lecture 17 Flow Analysis flow analysis in prolog; applications of flow analysis

Lecture 17 Flow Analysis flow analysis in prolog; applications of flow analysis. Ras Bodik Shaon Barman Thibaud Hottelier. Hack Your Language! CS164 : Introduction to Programming Languages and Compilers, Spring 2012 UC Berkeley. Today. Static program analysis what is it and why do it

lindsay
Download Presentation

Lecture 17 Flow Analysis flow analysis in prolog; applications of flow analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Lecture 17Flow Analysisflow analysis in prolog; applications of flow analysis

    Ras Bodik Shaon BarmanThibaud Hottelier Hack Your Language! CS164: Introduction to Programming Languages and Compilers, Spring 2012UC Berkeley
  2. Today Static program analysis what is it and why do it Points-to analysis static analysis for understanding how pointer values flow Andersen’s algorithm via deduction Andersen’s algorithm in Prolog just four lines Andersen’s algorithm via CYK parsing (optional) CFL-reachability
  3. Static program analysis Answers questions about program properties related to static type inference Static analysis == at compile time that is, prior to seeing the actual input hence, the answer must be correct for all inputs Sample program properties: Does var x have a constant value (for all inputs)? Does foo() return a table (whenever called, on all inputs)?
  4. Motivation for static program analysis (1) Optimize the program. Ex: replace x[i] with x[1] if we know that iis always 1. Constant propagation i = 2 … i = i+2 … if (…) { …} … x[i] = x[i-1]
  5. Motivation for static program analysis (2) Find potential security vulnerabilities Ex: in a server program, can a value flow from POST (untrusted, tainted source) to SQL interpreter (trusted sink) without passing through cgi.escape (a sanitizer)? This is taint analysis. Can be dynamically or static. Dynamic: mark values with a tainted bit. Sanitization clears the bit. An assertion checks that tainted values do not reach the interpreter. http://www.pythonsecurity.org/wiki/taintmode/ Static: a compile-time variant of this analysis. Proves that no input can ever make a tainted value flow to trusted sink.
  6. Motivation for static program analysis (3) Optimization of virtual calls in Java: virtual calls are costly, due to method dispatch Idea: Determine the target function of the call statically. If we can prove that the call has a single target, it is safe to rewrite the virtual call so that it calls the target directly. How to analyze whether a call has this property? Based on declared (static) types of pointer variables: Foo a = …; a.f() // a could call Foo::f or Bar::f. Cant’ tell from def of a By analyzing what values flow to a=…. That is, we try to compute the dynamic type of a more precisely than is given by the definition “Foo a”.
  7. Example class A { void foo() {…} } class B extends A { void foo() {…} } void bar(A a) { a.foo() } // can we optimize this call? B myB = new B(); A myA = myB; bar(myA); Declared type of a permits a.foo() to call both A::foo and B::foo. Yet we know only B::foo is the target, which allows optimization. What program property would reveal that the optimization is possible?
  8. Client 2: Verification of casts In Java, casts are checked at run time type system not expressive enough to check them statically although Java generics help somewhat The anatomy of a cast check: (Foo) e translates to if ( dynamic_type_of(e) not compatible with Foo ) throw ClassCast Exception t1 compatible with t2: t1 = t2 or t1 subclass of t2 Goal: prove that no exception will happen at runtime Why do this? The exception prevents any security holes, no? Such static verification useful to catch bugs (Mars Rover).
  9. Example class SimpleContainer { Object a; void put (Object o) { a=o; } Object get() { return a; } } SimpleContainer c1 = new SimpleContainer(); SimpleContainer c2 = new SimpleContainer(); c1.put(new Foo()); c2.put(“Hello”); Foo myFoo = (Foo) c1.get(); // verify that cast does not fail Note: analysis must distinguish containers c1 and c2. otherwise c1 will appear to contain string objects What property will lead to desired verification?
  10. Motivation for static program analysis (4) Compile 164 into efficient code If p always refers to tables that contains fields f1 and f2, we can represent the table as a struct and compile p[“f2”] into an (efficient) instruction “load from address in p + 4 bytes”. The analysis Determine at compile time what fields the object may ever contain at run time. A conservative rule (conservative=sufficient but not necessary): Compute, at compile time: the set of fields are added to the table using stmt e.ID=e the table’s fields must not be written or read through operator e[e] (only through e.ID)
  11. Discussion Why is e[e] dangerous? Consider: p[read_input_string()]=… creates a field whose name is unknown statically
  12. Example (JavaScript) var p = new Foo; // line 1 var r = p.field; var s = {}; s[r.f] = p; var t = s[input()]; t.g = … Consider the Foo objects created in line 1: Can we determine at compile time what fields these objects will contain during their lifetime (for any input)? If these objects are not accessed via e[e], then we can compute (a superset of) these fields. Can we tell if this program access Foo’s via e[e]?
  13. Static analysis must be conservative When unsure, the analysis must answer such that it does not mislead the client of the analysis. Err on the side of caution. Say, never optimize the program such that it outputs a different value. Several ways an analysis can be unsure: Property holds on some but not all execution paths. Property holds on some but not all inputs.
  14. Misleading the client: Constant propagation: if x is not always a constant but is claimed to be so by the analysis to the client (the optimizer), this would lead to optimization that changes the semantics of the program. The optimizer broke the program. Taintednessanalysis: Saying that a tainted value cannot flow may lead to missing a bug by the security engineer during program review. Yes, we want find to find all taintendness bugs, even if the analysis reports many false positives (ie many warnings are not bugs).
  15. What analysis that can serve these clients? Is there a program property useful to these clients? Yes. We want to understand how references “flow” References (pointer values): how are they copied from variable to variable? Flow from creation of an object to its uses that is, flow from new Foo to myFoo.f Note: the pointer may flow via the heap that is, a pointer may be stored in an object’s field ... and later read from this field
  16. Assume Java For now, assume we’re analyzing Java thanks to class defs, fields of objects are known statically (also, assume the Java program does not use reflection)
  17. Flow analysis as a constant propagation Initially we’ll only handle new and assignments p=r: if (…) p = new T1() else p = new T2() r = p r.f()// what are possible dynamic types of r?
  18. Flow analysis as a constant propagation We (conceptually) translate the program to if (…) p = o1 else p = o2 r = p r.f()// what are possible symbolic constant values r?
  19. Abstract objects The oiconstants are called abstract objects an abstract object oi stands for any and all dynamic objects allocated at the allocation site with number i allocation site = a new expression each new expression is given a number i When the analysis says a variable p may have value o7 we know that p may point to any object allocated in the expression “new7 Foo”
  20. We now consider pointer dereferences x = new Obj(); // o1 z = new Obj(); // o2 w = x; y = x; y.f = z; v = w.f; To determine abstract objects that v reference, what new question do we need to answer? Can y and w point to same object?
  21. Keeping track of the heap state Heap state: what objects a variable may point to at a particular program point. Heap state may change at each statement Analyses often don’t track state at each point separately to save space, they collapse all program points into one consequently, they keep a single heap state This is called flow-insensitive analysis why? see next slide
  22. Flow-Insensitive Analysis Disregards the control flow of the program assumes that statements can execute in any order … … and any number of times Effectively, flow-insensitive analysis transforms this if (…) p = new T1(); else p = new T2(); r = p; p = r.f; into this control flow graph: p = new T1() r = p p = r.f p = new T2()
  23. Flow-Insensitive Analysis Motivation: there is a single program point, and hence a single “version” of program state Is flow-insensitive analysis sound? yes: each execution of the original program is preserved and thus will be analyzed and its effects reflected But it may be imprecise it adds executions not present in the original program it does not distinguish value of p at distinct pgm points
  24. Let’s develop the analysis! Canonical Stmts Java pointers give rise to complex expressions: ex: p.f().g.arr[i] = r.f.g(new Foo()).h Can we find a small set of canonical statements ie, the core language understood by the analysis we’ll desugar the rest of the program to these stmts We only need four canonical statements: p = new T() new p = r assign p = r.fgetfield p.f = r putfield
  25. Canonical Statements, discussion Complex statements can be canonized p.f.g = r.f → t1 = p.f t2 = r.f t1.g = t2 Can be done with a syntax-directed translation like translation to byte code in PA2
  26. Handling of method calls Issue 1: Arguments and return values: these are translated into assignments of the form p=r Example: Object foo(T x) { return x.f } r = new T; s = foo(r.g) is translated into foo_retval = x.f r = new T; s = foo_retval; x = r.g
  27. Handling of method calls Issue 2: targets of virtual calls call p.f() may call many possible methods to do the translation shown on previous slide, must determine what these targets are Suggest two simple methods:
  28. Handling of arrays We collapse all array elements into one element this array element will be represented by a field arr ex: p.g[i] = r becomes p.g.arr = r
  29. Andersen’s Algorithm For flow-insensitive flow analysis: Goal: compute two binary relations of interest: x pointsTo o: holds when x may point to abstract object o o flowsTo x: holds when abstract object o may flow to x These relations are inverses of each other x pointsTo o <==>o flowsTo x
  30. These two relations support our clients These relations allows determining: target methods of virtual calls verification of casts how JavaScript objects are used For 3) we need the flowsTo relation For 1) and 2) we need the x pointsToo relation
  31. Inference rule (1) p = newi T() oinew p oinew p →oiflowsTo p
  32. Inference rule (2) p = r r assign p oiflowsTo r r assign p →oiflowsTop
  33. Inference rule (3) p.f= a a pf(f) p b = r.fr gf(f) b oiflowsTo a  a pf(f) p  p alias r  r gf(f) b →oiflowsTob
  34. Inference rule (4) it remains to define x alias y (x and y may point to same object): oiflowsTo x oiflowsTo y →x alias y
  35. Prolog program for Andersen algorithm new(o1,x). % x=new_1 Foo() new(o2,z). % z=new_2 Bar() assign(x,y). % y=x assign(x,w). % w=x pf(z,y,f). % y.f=z gf(w,v,f). % v=w.f flowsTo(O,X) :- new(O,X). flowsTo(O,X) :- assign(Y,X), flowsTo(O,Y). flowsTo(O,X) :- pf(Y,P,F), gf(R,X,F), aliasP,R), flowsTo(O,Y). alias(X,Y) :- flowsTo(O,X), flowsTo(O,Y).
  36. How to use the result of the analysis? When the analysis infers o flowsToy, what did we prove? nothing useful, usually, since oflowsToy does not imply that there is a program input for which o will definitely flow to y. The useful result is when the analysis can’t infer o flowsTo y then we have proved that o cannot flow to y for any input this is useful information! it may lead to better optimization, verification, compilation Same arguments apply to alias, pointsTo relations and other static analyses in general
  37. Inference Example (1) The program: x = new Foo(); // o1 z = new Bar(); // o2 w = x; y = x; y.f = z; v = w.f;
  38. Inference Example (2): The program is converted to six facts: o1 new x o2 new z x assign w xassign y z pf(f)y wgf(f) v
  39. Inference Example (3), infering facts o1 new x o2 new z x assign w x assign y z pf(f)y wgf(f) v The inference: o1 new x →o1flowsTox o2 new z →o2flowsToz o1flowsTox x assign w →o1flowsTow o1flowsTox x assign y →o1flowsToy o1flowsToy  o1flowsTow →y alias w o2flowsToz  z pf(f)y y alias wwgf(f) v →o2flowsTov ...
  40. Example: visualizing Prolog deductions o1 new x assign assign w y gf[f] pf[f] new v z o2
  41. Example, deriving the relations o1 new x assign assign w y gf[f] pf[f] new v z o2
  42. Example (4): Notes: inference must continue until no new facts can be derived only then we know we have performed sound analysis Conclusions from our example inference: we have inferred o2flowsTov we have NOT inferred o1flowsTov hence we know v will point only to instances of Bar (assuming the example contains the whole program) thus casts (Bar) v will succeed similarly, calls v.f() are optimizable
  43. “Parsing the graph” Visualization of inferences on slides 41 and 42 parses the strings in the “graph of binary facts” using the CYK algorithm (Lecture 8) Details on this style of inference are in the rest of the slide, under CFL-reachability (optional material)
  44. Adaptation for JavaScript Need to handle more language constructs: property read e1[e2] property write e1[e2] = e3 assume that e2 can return any value, and the analysis does not analyze the value Extensions to the algorithm: analysis must determine whether an object might appear as e1 in e1[e2] = e3 if yes, we must conservatively assume that we don’t know objects fields more similar rules are needed … 46
  45. Summary Determine run-time properties of programs statically example property: “is variable x a constant?” Statically: without running the program it means that we don’t know the inputs and thus must consider all possible program executions We want sound analysis: err on the side of caution. allowed to say x is not a constant when it is not allowed to say x is a constant when it is not Static analysis has many clients optimization, verification, compilation
  46. The technique Flow-insensitive analysis: collapse into one all program points (ie, stmt entry and exits) reduces the amount of analysis state to maintain reduces precision, too, of course Transform this program if (…) p = new T1(); else p = new T2(); r = p; p = r.f; into this one: p = new T1() r = p p = r.f p = new T2()
  47. Andersen’s algorithm Deduces the flowsTo relation from program statements statements are facts analysis is a set of inference rules flowsTo relation is a set of facts inferred with analysis rules Statement facts: we’ll write them as x predicateName y p = newi T() oinew p p = r r assign p p = r.fr gf(f) p p.f = r r pf(f) p
  48. CFL-Reachability

    deduction via parsingof a graph
  49. Inference via graph reachability Prolog’s search is too general and expensive. may in general backtrack (exponential time) Can we replace it with a simpler inference algorithm? possible when our inference rules have special form We will do this with CFL-rechability it’s a generalized graph reachability
  50. (Plain) graph reachability Reachability Def.: Node x is reachable from a node y in a directed graph G if there is a path p from y to x. How to compute reachability? depth-first search, complexity O(N+E)
  51. e ) e ] s t [ ( e [ e ] ] Context-Free-Language-Reachability CFL-Reachability Def.: Node x is L-reachable from a node y in a directed labeled graph G if there is a path p from y to x, and path p is labeled with a string from a context free language L. The context-free language L: matched →matched matched | ( matched ) | [ matched ] | e |  Is t reachable from s according to the language L?
  52. Computing CFL-reachability Given a labeled directed graph P and a grammar G with a start nonterminal S, we want to compute whether x is S-reachable from y for all pairs of nodes x,y or for a particular x and all y or for a given pair of nodes x,y We can compute CFL-reachability with CYK parser x is S-reachable from y if CYK adds an S-labeled edge from y to x O(N3) time
  53. Convert inference rules to a grammar The inference rules ancestor(P,C) :- parentof(P,C). ancestor(A,C) :- ancestor(A,P), parentof(P,C). Language over the alphabet of edge labels ANCESTOR ::= parentof | ANCESTOR parentof Notes: initial facts are terminals (perentof) derived facts are non-terminals (ANCESTOR)
  54. So, which rules can be converted to CFL-reachability? ANCESTOR ::=parentof | ANCESTOR parentof Is “son” ANCESTOR-reachable from “grandma”? grandma mom me son parentof parentof parentof ANCESTOR ANCESTOR ANCESTOR
  55. What rules can we convert to CFL-rechability? Let’s add a rule for SIBLING: ANCESTOR ::=parentof | ANCESTOR parentof SIBLING ::=??? We want to ask whether “bro” is SIBLING-reachable from “me”. grandma mom me son parentof parentof parentof parentof bro
  56. Conditions for conversion to CFL-rechability Not all inference rules can be converted Rules must form a “chain program” Each rule must be of the form: foo(A,D) :- bar(A,B), baz(B,C), baf(C,D) Ancestor rules have this form ancestor(A,C) :- ancestor(A,P), parentof(P,C). But the Sibling rules cannot be written in chain form why not? think about it also from the CFL-reachability angle no path from x to its sibling exists, so no SIBLING-path exists no matter how you define the SIBLING grammar
  57. Andersen’s Algorithm with Chain Program

    converts the analysis into a graph parsing problem
  58. Back to Andersen’s analysis Rules in logic programming form: flowsTo(O,X) :- new(O,X). flowsTo(O,X) :- flowsTo(O,Y), assign(Y,X). flowsTo(O,X) :- flowsTo(O,Y), pf(Y,P,F), alias(P,R), gf(R,X,F). alias(X,Y) :- flowsTo(O,X), flowsTo(O,Y). Problem: some predicates are not binary
  59. Andersen’s algorithm inference rules Translate to binary form put field name into predicate name, must replicate the third rule for each field in the program flowsTo(O,X) :- new(O,X). flowsTo(O,X) :- flowsTo(O,Y), assign(Y,X). flowsTo(O,X) :- flowsTo(O,Y), pf[F](Y,P), alias(P,R), gf[F](R,X). alias(X,Y) :- flowsTo(O,X), flowsTo(O,Y).
  60. Andersen’s algorithm inference rules Now, which of these rules have the chain form? flowsTo(O,X) :- new(O,X). yes flowsTo(O,X) :- flowsTo(O,Y), assign(Y,X). yes flowsTo(O,X) :- flowsTo(O,Y), pf[F](Y,P), alias(P,R), gf[F](R,X). yes alias(X,Y) :- flowsTo(O,X), flowsTo(O,Y). no
  61. Making alias a chain rule We can easily make alias a chain rule with pointsTo. Recall: flowsTo(O,X) :- pointsTo(X,O) pointsTo(X,O) :- flowsTo(O,X) Hence alias(X,Y) :- pointsTo(X,O), flowsTo(O,Y). If we could derive chain rules for pointsTo, we would be done. Let’s do that.
  62. Idea: add terminal edges also in opposite direction For each edge o new x, add edge x new-1 o same for other terminal edges Rules for pointsTo will refer to the inverted edges but otherwise these rules are analogous to flowsTo What it means for CFL reachability? there exists a path from o to x labeled with s  L(flowsTo)  there exists a path from x to o labeled with s’L(pointsTo).
  63. Inference rules for pointsTo p = newi T() oinew p pnew-1oi oi new p →oiflowsTo p Rule 1 p new-1oi→ppointsTooiRule 5 p = r r assign p p assign-1 r oiflowsTo r andr assign p → oiflowsTo p Rule 2 p assign-1 r andr pointsTooi→ppointsTooiRule 6
  64. Inference rules for pointsTo (Part 2) We can now write alias as a chain rule. p.f = a a pf(f) p p pf(f)-1a b = r.fr gf(f) b b gf(f)-1 r oiflowsTo a a pf(f) p p alias r r gf(f) b →oiflowsTo b b gf(f)-1 r r alias p p pf(f)-1 a a flowsTooi→bpointsTooi Rules 3, 7 Both flowsTo and pointsTo use the same alias rule: x pointsTooioiflowsTo y →x alias y Rule 8
  65. The reachability language All rules are chain rules now directly yield a CFG for flowsTo, pointsTo via CFL-reachability : flowsTo→new flowsTo→ flowsTo assign flowsTo→flowsTopf[f] alias gf[f] pointsTo→new-1 pointsTo→assign-1pointsTo pointsTo→gf[f]-1 alias pf[f]-1pointsTo alias →pointsToflowsTo 67
  66. Example: computing pointsTo-, flowsTo-reachability Inverse terminal edges not shown, for clarity. o1 new x assign assign w y gf[f] pf[f] new v z o2
  67. Summary (Andersen via CFL-Reachability) The pointsTo relation can be computed efficiently with an O(N3) graph algorithm Surprising problems can be reduced to parsing parsing of graphs, that is
  68. CFL-Reachability: Notes The context-free language acts as a filter filters out paths that don’t follow the language We used the filter to model program semantics we filter out those pointer flows that cannot actually happen What do we mean by that? consider computing x pointsTo o with “plain” reachability plain = ignore edge labels, just check if a path from x to o exists is this analysis sound? yes, we won’t miss anything we compute a superset of pointsTo relation based on CFL-reachability but we added infeasible flows, example: wrt plain reachability, pointer stored in p.f can be read from p.g
More Related